Re: Newbie question about db normalization theory: redundant keys OK?

On Dec 14, 12:30 pm, -CELKO- <jcelko..._at_earthlink.net> wrote:
> >> If teachers are permitted to change their names, then names are not appropriate keys for teachers. <<
>
> Agreed; this was a skeleton table to demonstrate to the OP that tables
> can have redundant or even overlapping keys to preserve data
> integrity. Tony just gets easily side tracked and I am afraid I tend
> to fall for it myself. Sorry for the thread drift ..
>
> Instead of a teacher's name, we can tattoo a permanent bar code on the
> foreheads of the faculty, using their tax identification number :)

Celko what do you think of 1NF, 2NF, 3NF as programming exercises? (see here for my understanding of these concepts: http://www.utexas.edu/its-archive/windows/database/datamodeling/rm/rm7.html ) Obviously for 'cascades' of UPDATE changes and the like, 3NF is the easiest to code, but getting a dB into 3NF form if it's in 2NF or 1NF form is a bit of work, agreed? The U of Texas example was easy to follow but in the real world I doubt it's so cut and dry.

Keep in mind I have a few days experience in this matter, but I think I'm on the right track with the above statement.

RL Received on Fri Dec 14 2007 - 22:51:21 CET